Bach used the Deceptive Cadence as early as his Passacaglia and Fugue in C minor, BWV 582, which dates to the early 18th century. Are there earlier uses? A definitive answer may be impossible, but I am interested in how far back its usage can be traced.
The earliest two references I can find of what are now known as deceptive cadences originate with Josquin des Prez's Missa Una musque de Buscaya (listed without a date on Wikipedia, which only fleetingly mentions it under a different spelling, suggesting that his authorship is doubtful) and Francesco Spinacino's arrangement of Fortuna dun gran tempo. I was not able to find particular dates for either of these works, but based on the known dates of their authors, this places deceptive cadences at least as far back as the early 1520s, if not as far back as the late 15th century.
The article I found these in itself unfortunately requires access to JSTOR, but below are the two relevant parts of the article.
Josquin based this mass on the popular tune Une mousse de Biscaye. The several phrases of this Hypo-Lydian melody cadence on g, f, g, b-flat, g, f, b-flat. Within the mass however, the bassus frequently adds a third or a fifth beneath these cadence tones--recognizable in the tenor at the several closings within the different parts of the mass--creating (in modern terminology) a 'deceptive cadence'. To the cadence tone f in the tenor the bassus always adds another tone--d or B-flat--disturbing the listener's unambiguous rientation towards a final cadence in the fa-mode
Within the two periods bordered by the three Lydian cadences just mentioned (mm. 10-17 and 17-25), the avoidance by Spinacino of the b-natural leading tone weakens the temporary orientation towards C in favour of the following cadences (mm. 144-15', 22242 3 ) which, in turn, are disturbed by means of a 'deceptive cadence', the result of the contra's introduction of the second and third melody segment.